Hydrobiologia (2007) 589:235–248 DOI 10.1007/s10750-007-0733-3

PRIMARY RESEARCH PAPER

Growth and resource allocation of indica and Schoenoplectus validus as affected by interspecific competition and nutrient availability

Zhenhua Zhang Æ Zed Rengel Æ Kathy Meney

Received: 12 November 2006 / Revised: 12 April 2007 / Accepted: 18 April 2007 / Published online: 13 June 2007 Springer Science+Business Media B.V. 2007

Abstract Nutrient availability and interspecific in the high than the low nutrient treatment. The competition may affect emergent wetland total biomass for C. indica in mixture increased growth and resource allocation in constructed significantly in the high nutrient treatment, but wetland. A glasshouse study was conducted to that for S. validus was significantly lower in investigate the influence of nutrient and mixture mixture than in monoculture. Relative yield between Linn and Schoenoplectus (RY) indicated that there was significant inter- validus (Vahl) A. Lo¨ ve & D. Lo¨ ve on their specific competition between S. validus and C. in- growth and resource allocation in the wetland dica in mixtures, with C. indica being the superior microcosms, using simulated secondary-treated competitor. The growth of S. validus was signif- municipal wastewater effluent with either low icantly inhibited by the presence of C. indica in (17.5 mg N and 10 mg P l–1) or high (35 mg N and their mixture. Compared with monoculture, 20 mg P l–1) nutrient concentrations. After S. validus in mixture had significantly higher 65 days, the high nutrient treatment stimulated percentages of root biomass and allocations of plant growth and resulted in allocation of more N and P to roots, whereas C. indica was not resources to the above-ground tissues compared significantly affected by mixture. The results to below-ground ones. The concentrations of N suggested that the growth and resource allocation and P in the plant tissues (except P in above- of C. indica and S. validus could be altered by ground tissues) were significantly higher, whereas nutrient availability and interspecific competition N and P use efficiencies were significantly lower in constructed wetlands.

Keywords Canna indica Growth Interspecific Handling editor: S. M. Thomaz competition Nutrient Resource allocation Schoenoplectus validus Z. Zhang Z. Rengel (&) Soil Science and Plant Nutrition, School of Earth and Geographic Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Introduction Australia e-mail: [email protected] The use of constructed wetlands for tertiary water purification of municipal wastewater has been K. Meney Syrinx Environmental Pty Ltd, 12 Monger St, Perth, received an increasing attention in recent years. WA 6000, Australia Constructed wetlands offer effective and reliable

123 236 Hydrobiologia (2007) 589:235–248 treatment of wastewater in a simple and inexpen- ture) systems are more susceptible to plant death sive manner (Sundaravadivel & Vigneswaran, due to predation or disease. Therefore, it is 2001). generally assumed that multiple plant (mixture) Wetland contribute to the function of and native plant systems are more resilient than wetland systems as habitat providers, bioengi- monocultures (EPA, 2000). It is well known that neers to promote sedimentation and even disper- positive, negative or indifferent interrelationship sion of flow, and erosion control and nutrient/ may occur between plants of different species. pollutant transformers (Kadlec & Knight, 1996). The results of such competition might cause the This latter attribute relates to the rapid growth preferential establishment and growth of certain rates of species in resource-rich environments, species, and/or the suppression and extinction of and ability to concentrate luxury amounts of other species (Agami & Reddy, 1990). In recent nutrients in their above- and below-ground bio- years, several studies have been reported on the mass. The partitioning of nutrients between interspecific competition for emergent wetland shoots and roots/ varies between species species. For example, Wetzel and van der Valk and seasons. In small, lightly loaded wastewater (1998) found that Phalaris arundinacea is an treatment systems, plant uptake can be the inherently better competitor than Carex stricta principal form of nutrient removal, often account- or Typha latifolia. Coleman et al. (2001) observed ing for up to 80% of the nutrient pool (Kadlec & that Typha latifolia was the superior competitor Knight 1996). Plants are also the principal nutri- among the three-species (Juncus effusus, Typha ent sinks during the initial years of establishment latifolia and Scirpus cyperinus) mixture in small- in the wetland. The removal of plant material scale constructed wetlands. In plant mixtures through a harvest reduces the potential for consisting of Carex flava, Centaurea angustifolia, biologically-assimilated nutrients being remobi- Lycopus europaeus and Selinum carcifolia grown lised into the wetland system. Harvesting can also in the sand culture with different N:P supply encourage large nutrient uptake by the plants ratios and different total supplies of N and P, during the rapid growth and recovery of the Lycopus europaeus performed best at low and harvested plants. Therefore, the function of plants intermediate N:P ratios, and Carex flava at high in nutrient stripping is a dynamic one, and N:P ratio (Gu¨ sewell & Bollens, 2003). However, requires an understanding of plant tolerances few studies have investigated the competitive and nutrient requirements to optimize perfor- impact between species with different growth mance in terms of uptake. forms or significantly different morphologies. The concentrations of nutrients (N and P) in Schoenoplectus validus (Vahl) A. Lo¨ ve & D. the effluents of the wastewater and loading rate to Lo¨ ve and closely related rush species S. lacustris, the constructed wetlands vary at different times S. acutus and S. californicus have been used during the year and for different wastewater widely in constructed wetlands around the world treatment plants. On the other side, nutrient (Tanner, 2001). S. validus is a tall, perennial, availability could affect plant growth responses herbaceous rush. The are reduced to and resource allocations. Plant not only grow at a sheaths at the base of the stems. It usually occurs slow rate at low nutrient supply compared with in poorly drained soil, grows better in saline high nutrient supply, but also increase their conditions than in fresh water and tolerates a biomass allocation to roots (Poorter & Nagel, wide range of salinity. Similarly, ornamental 2000) and reduce the nutrient concentrations in species, Canna indica Linn, has been used to their biomass (Aerts & Chapin, 2000). treat (i) septic tank effluent in a simulated Improvements in plant selection and cultiva- vertical-flow constructed wetlands (Zhu et al., tion might make the constructed wetlands more 2004); (ii) domestic wastewater in a medium-scale efficient for nutrient removal from the wastewa- vertical/reverse-vertical flow constructed wetland ter. It is not clear if it is desirable to maintain a (Yue et al., 2004); and (iii) municipal wastewater single plant species, or a mix of plant species, in in a full-scale subsurface-flow constructed wet- constructed wetland. But single plant (monocul- land (Shi et al., 2004). C. indica is an upright

123 Hydrobiologia (2007) 589:235–248 237 perennial rhizomatous herb. The leaves are ellip- (0.39 · 0.29 · 0.30 m) with a hole fitted with a tic and rather fleshy. It is a native plant of tropical plastic tube close to the base to facilitate drainage America, and very popular ornamental plant of water. A mesh covering the hole was fastened throughout the tropical and subtropical region on the inside of the microcosm to prevent loss of around the world, but grows in thickets, crowding sand during the water drainage. Approximately out other plants and difficult to remove due to its 25 kg of washed river sand (<4 mm diameter) was spread by rhizomes. added to each container, giving a sand depth of Although the two plant species and their about 0.15 m. closely related species have been selected and planted in the various constructed wetlands for Experimental setup the improvement of water quality and landscape restoration under mono- or mixed-culture con- The experiment was conducted in a phytotron at ditions, due to their relative high nutrient the University of Western Australia with con- removal efficiency and aesthetical value (Cal- trolled day/night temperatures of 25/20C under heiros et al., 2007; Fu et al., 2006; Grosse et al., natural light conditions from the beginning of 2001; Wu and Ding, 2006; Wu et al., 2006), no June to the later September. A complete ran- research has been done on dealing with inter- domized block factorial design (four plant culture specific competition between C. indica and treatments · two nutrient concentration treat- S. validus. The objective of this study was to ments) with three replicates was employed. investigate the influence of nutrient availability At the establishment of plant treatment, the and mixture between C. indica and S. validus on four plant culture treatments with nine replicates their growth and resource allocation using were: (a) control (with no plants added); (b) simulated secondary-treated municipal wastewa- monoculture of C. indica; (c) monoculture of ter in the wetland microcosms. S. validus; and (d) mixture between C. indica and S. validus. The seedlings of C. indica and S. validus were purchased from a local nursery. The healthy plants of relatively similar size were Materials and methods transplanted into the microcosms. Each micro- cosm contained six plants in two row of three: six Concentrations of N and P in secondary- C. indica, six S. validus or three of each species treated municipal wastewater effluent (Fig. 1). At the planting time, seedlings of C. indica were approximately 8 cm tall with 1–2 The data on quality of discharge water from leaves, and S. validus were approximately 10 cm secondary-treated municipal wastewater effluent tall with 4–6 ramifications. collected at Subiaco Waste Water Treatment In order to minimize variability in the exper- Plant in Shenton Park, Perth, Western Australia iment, a simulated nutrient solution was used in from July 2002 to November 2004 were obtained the microcosms at the average concentrations of from Water Corporation, Western Australia. The N and P similar to the secondary-treated waste- yearly ranges of the total N and P concentrations water at Subiaco Waste Water Treatment Plant as in the wastewater effluent were from 8.2 to 28 mg described above. The solution contained 17.5 mg Nl–1 and 5.6 to 17 mg P l–1; and the average Nl–1 (1:1 NH –N and NO –N) and 10 mg P l–1, concentrations between 1 November 2003 to 31 4 3 and other macro- and micronutrients (mg l–1): 13 October 2004 were 17 mg N l–1 and 10 mg P L–1. K, 10 Ca, 5 Mg, 7 S, 0.04 Zn, 0.01 Cu, 0.34 Fe, 0.05 Mn, 0.26 B and 0.05 Mo. Microcosms were Wetland microcosms filled with the nutrient solution to achieve the water depth of 0.02 m above the sand surface and Thirty-six microcosms were used as experimental renewed weekly. units in this study. A wetland microcosm was After plants were grown in the microcosms for established in 33-l plastic containers 50 days, three replicates of each plant culture

123 238 Hydrobiologia (2007) 589:235–248

Fig. 1 The vertical view of the design plant treatment in the microcosms. (a) control (with no plants added); (b) monoculture of C. indica;(c) monoculture of S. validus; and (d) mixture between C. indica and S. validus treatment (including control) were harvested, and To estimate above-ground biomass of plants plant (shoot, and root) and substrate prior to the destructive sampling, for S. validus, samples were taken. The remaining microcosms regression relationship between total shoot length were initialized with two nutrient treatments each and dry above-ground biomass were developed at in three replicates: (1) low level containing the end of the experiment. For C. indica,no nutrients as in the starting solution described significant regression between shoot number and above, and (2) high level with double concentra- above-ground biomass was detected, so that shoot tions of N and P. The concentrations of other number was not included in the regression. These nutrients were kept the same as in the starting equations were used to estimate above-ground solution described above, except double K. biomass from weekly shoot height and number Microcosms were filled with 10 l either low or measurements. The regression equation for each high nutrient solution to achieve the water depth species was: of about 0.08 m above the sand surface. Each S. validus microcosm was manually drained through tube positioned at the bottom of each container and Ln above ground biomassðÞ g refilled every 7 days. ¼ 6.399 þ 0.699 Ln length of total shootðÞ cm , ÀÁ r2 ¼ 0.993, P \ 0.001 Sampling and measurements C. indica Total shoot numbers and shoot heights in each microcosm were measured at approximately Ln above ground biomassðÞ g weekly intervals from the day after imposing ¼ 1.281 þ 0.964 Ln shoot heightðÞ cm , nutrient treatments until the plants were har- ÀÁ 2 vested at the end of the experiment (65 days r ¼ 0.968, P \ 0.001 later). The shoot height was measured from the base of the plants to the top of the longest leaves The plants were harvested after 65 days of for C. indica or the top of culms for S. validus. nutrient and plant culture treatments. The shoots

123 Hydrobiologia (2007) 589:235–248 239 were cut at the sand surface, and their bases were number = (Ln final shoot number–Ln initial washed to remove any adhering sediments. Each shoot number)/days (Tylova-Munzarova et al., microcosm was then excavated and hand sorted 2005). into above-ground plant parts (stems and leaves, The competitive effects of species were exam- inflorescences and flowering stems), rhizomes ined by calculating the relative yield of individ- (including stem base) and roots. The plants that uals of a species when grown with another species had grown in mixture were separated to species. (interspecific competition) compared to the rela- The roots were separated from the sand by tive yield of the species grown alone (intraspecific washing away the sand using tap water and competition) while maintaining the same overall collecting roots onto the mesh. All plant samples density. Relative yield (RY) of above-ground were dried to constant weight at 70C for 5 days biomass was calculated as: in a forced-air oven, weighed and ground to pass a RY of species A = (yield of species A in 0.75-mm mesh. mixture)/(yield of species A in monoculture) Total nitrogen in the samples of plant tissues (Harper, 1977). were determined by the Dumas combustion All statistical tests were performed using SPSS method using an automated CN analyzer (LECO version 10 for windows. A set of two-way ANO- CHN-1000, LECO Company, St Joseph, Michi- VA was used to determine significance of nutrient gan, USA). Total phosphorus in plant was deter- and plant treatment effects on plant biomass and mined by colorimetric with spectrophotometer characteristics, plant concentrations of N and P, HITACHI U-1100 using the vanado-molybdate and resource allocation (the percentage after log method after digesting material in mixture of transformation). Least significant difference concentrated nitric and perchloric acids (Bassett (LSD) was applied to test for significance et al., 1978). between treatment means.

Calculations and statistical analysis

All data for calculation and analysis were Results obtained from the beginning of imposing nutrient treatments to the end of the experiment (65 days Plant biomass of plant growth). Dry biomass production was estimated from the total biomass of each micro- Significant differences in the aboveground and cosm divided by the area of the microcosm. The total biomass were observed, but not in the biomass allocation was characterized using the belowground, rhizome and root biomass between ratio of root-supported tissue (above-ground, the nutrient treatments and significant differences rhizomes) to root biomass (S/R) and the ratio of in the biomass of various plant parts were observed above-ground to below-ground biomass (A/B) among the plant treatments. The significant inter- (Lorenzen et al., 2001). Resource allocation ratio actions between nutrient and plant treatments was defined as the ratio of a certain tissue biomass were observed in the aboveground and total or N, P content to total plant biomass or total biomass (Table 1). After 65 days of nutrient and plant N, P content. The allocation into above- plant treatments, the total biomass for C. indica in ground, rhizomes and roots were calculated as the mixture was significantly higher in the high nutri- ratio between the biomass or nutrient content of ent treatment than in the low nutrient treatment, the fraction and the total biomass or nutrient but that for S. validus was not significantly affected content of the plants. Nutrient use efficiency was by the nutrient treatments. The above-ground calculated as the total dry biomass divided by biomass for C. indica in both monoculture and total N or P content. The following biometric mixture, and that for S. validus in monoculture characteristics were estimated: relative shoot were significantly increased by the high nutrient growth rate = (Ln final shoot length–Ln initial treatments. The below-ground biomass for C. in- shoot length)/days, and relative increase in shoot dica was significantly higher, whereas that for

123 240 Hydrobiologia (2007) 589:235–248

Table 1 Mean (±SE, n = 3) dry weight (kg m–2) in different parts of plants influenced by nutrient and plant treatments after 65 days of plant growth, with analysis of variance Treatment Total Above-ground Below-ground Rhizome Root

Low nutrient Mono-C. indica 1.77 ± 0.10bc 0.48 ± 0.06d 1.29 ± 0.06c 1.16 ± 0.06c 0.13 ± 0.01c Mixed-C. indica 2.23 ± 0.10b 0.59 ± 0.04cd 1.64 ± 0.06b 1.48 ± 0.03b 0.16 ± 0.05c Mono-S. validus 1.60 ± 0.10c 0.94 ± 0.02bd 0.66 ± 0.12d 0.40 ± 0.09de 0.26 ± 0.03a Mixed-S. validus 0.74 ± 0.07d 0.36 ± 0.05d 0.38 ± 0.02e 0.18 ± 0.02e 0.20 ± 0.01abc High nutrient Mono-C. indica 2.18 ± 0.15b 0.86 ± 0.08bc 1.32 ± 0.07c 1.17 ± 0.07c 0.15 ± 0.01c Mixed-C. indica 3.51 ± 0.35a 1.59 ± 0.19a 1.92 ± 0.20a 1.77 ± 0.19a 0.15 ± 0.01c Mono-S. validus 2.09 ± 0.16b 1.34 ± 0.11a 0.75 ± 0.15d 0.51 ± 0.12d 0.24 ± 0.05ab Mixed-S. validus 0.97 ± 0.08d 0.59 ± 0.05cd 0.39 ± 0.04e 0.22 ± 0.03e 0.17 ± 0.02bc Source of variation Nutrient treatment *** *** NS NS NS Plant treatment *** *** *** *** ** Nutrient · plant * ** NS NS NS Means with different letters within columns are significantly different based on LSD (P < 0.05) NS: not significant. * Significant at P < 0.05. ** Significant at P < 0.01. *** Significant at P < 0.001

S. validus was significantly lower in mixture than in higher percentage of above-ground biomass, rel- monoculture treatment (Table 1). ative rate of shoot growth and relative increase in Species mixture remarkably influenced above- numbers of shoots in the high than the low ground plant growth before imposing nutrient treatments. For both species, these early effects 1.2 followed the same trends as those at harvesting Low nutrient time. After imposing nutrient treatments, the High nutrient relative growth of C. indica in mixture increased 0.8 in the high nutrient treatment, but the effect disappeared after day 40, whereas the relative S. validus 0.4 growth of S. validus in mixture was not enhanced Relative yield by the high nutrient treatment. Those effects in 65 days of plant growth were shown using the 0.0 relative yield (RY) of above-ground biomass for 2.0 C. indica and S. validus under the low and high 1.6 nutrient treatments (Fig. 2). The relative yield 1.2 (RY) indicated that there was significant inter- C. indica Low nutrient specific competition between S. validus and C. in- 0.8 High nutrient dica in mixture, with C. indica being the superior Relative yield 0.4 competitor. 0.0 0204060 Biomass allocation and biometric Days after imposing nutrient treatment (d) characteristics Fig. 2 Relative yield (RY) of above-ground biomass for C. indica and S. validus under the low and high nutrient Biomass allocation and plant characteristics (ex- treatments after imposing the nutrient treatments. The cept relative rate of shoot growth among the plant high nutrient treatment was imposed after transplanting treatments) were significantly affected by the plants of 50 days. A RY of 1 indicates that plants were the nutrient and plant treatments, but no significant same size in mixture as in monoculture. RY <1 indicates that plants were smaller in mixture than in monoculture interaction between nutrient and plant treatments and RY >1 indicates that plants were larger in mixture was detected (Table 2). Plants had significantly than in monoculture

123 Hydrobiologia (2007) 589:235–248 241

Table 2 Analysis of variance for biomass allocation to different tissues and biometric characteristics of plants between nutrient and plant treatments after 65 days of plant growth Variable Source of variability Nutrient treatment Plant treatment Interaction

Biomass allocation A/B ratio ** *** NS S/R ratio *** *** NS Aboveground percentage *** *** NS Rhizomes percentage ** *** NS Roots percentage *** *** NS Biometric characteristics Relative rate of shoot growth * NS NS Relative increase in shoot number ** *** NS NS: not significant. * Significant at P < 0.05. ** Significant at P < 0.01. *** Significant at P < 0.001 nutrient treatment. Compared with monoculture, nutrient treatments) were significantly affected S. validus in mixture had significantly lower A/B by nutrient and plant treatments, and significant ratio, percentage of above-ground biomass and interactions between nutrient and plant treat- relative increase in numbers of shoots under the ments were detected in N allocation to rhizome low nutrient treatment, and S/R ratio and relative and P allocation to above-ground organ and increase in numbers of shoots under the high rhizome (Table 5). Plants had significantly high- nutrient treatment, whereas C. indica was not er relative allocation of N and P to above- significantly affected by mixtures. Compared with ground organs in the high than the low nutrient C. indica, S. validus had more biomass allocated treatment. Compared with monoculture, into above-ground tissues (Table 3). S. validus in mixture had significantly higher relative allocation of N and P to roots (except P Concentrations of N and P in plant tissues allocation to root under the high nutrient treatment) and significantly lower N/P ratios in The concentrations of N and P in the plant tissues above-ground tissues and rhizomes, whereas were significantly influenced by nutrient and plant C. indica was not significantly affected by mix- treatments after 65 days of plant growth. The tures (Table 6). concentrations of N and P (except P in above- ground tissues) were significantly higher in the Nutrient use efficiency high nutrient than in the low nutrient treatment. The concentrations of N and P in plant tissues N and P use efficiency of C. indica and S. validus (except rhizome) were significantly different in were significantly affected by the nutrient treat- plant treatments. The concentrations of N and P ments, but not by the plant treatments. No in the roots of S. validus were significantly lower significant interaction was detected between in mixed-culture than in monoculture in the high nutrient and plant treatments (Table 7). N and nutrient treatment (Table 4). P use efficiency were significantly higher in the low nutrient treatment (86 g dry weight g–1 N and Nutrient allocation 474 g dry weight g–1 P, averaged over the plant treatments) than in the high nutrient treatment Plant nutrient (N and P) allocations and N/P (61 g dry weight g–1 N and 404 g dry weight g–1 P, ratios (except N/P ratios in roots between averaged over the plant treatments).

123 242 123

Table 3 Mean (±SE, n = 3) biomass allocation and biometric characteristics of plants influenced by nutrient and plant treatments after 65 days of plant growth Treatment Low nutrient High nutrient Mono- Mixed- Mono-S. validus Mixed- Mono-C. indica Mixed- Mono- Mixed- C. indica C. indica S. validus C. indica S. validus S. validus

Biomass allocation A/B ratio 0.5 ± 0.1d 0.5 ± 0.04d 1.5 ± 0.2b 1.0 ± 0.1c 0.7 ± 0.03cd 0.8 ± 0.1cd 2.0 ± 0.4a 1.6 ± 0.1ab S/R ratio 9.3 ± 0.4b 9.4 ± 1.3b 5.2 ± 0.3c 3.1 ± 0.3c 10.4 ± 0.7a 13.6 ± 0.9a 10.2 ± 2.4b 4.8 ± 0.1c Aboveground % 33 ± 2e 32 ± 0.3e 59 ± 4b 51 ± 2c 41 ± 1d 46 ± 2cd 66 ± 5a 61 ± 2a Rhizomes % 57 ± 2a 58 ± 2a 25 ± 3c 24 ± 0.3c 50 ± 0.3b 47 ± 2b 25 ± 4c 22 ± 2c Roots % 10 ± 0.4c 10 ± 2c 16 ± 1b 25 ± 2a 9 ± 0.4cd 7 ± 0.6d 10 ± 2c 17 ± 0.4b Biometric characteristics Relative rate of 0.023 ± 0.018c 0.030 ± 0.017bc 0.048 ± 0.023abc 0.037 ± 0.015abc 0.053 ± 0.008abc 0.074 ± 0.008a 0.068 ± 0.011ab 0.047 ± 0.005abc shoot growth (mm cm–1 d–1) Relative increase in 0.014 ± 0.001a 0.012 ± 0.002a 0.009 ± 0.001b 0.005 ± 0.001c 0.015 ± 0.001a 0.014 ± 0.001a 0.013 ± 0.001a 0.007 ± 0.002bc shoot number (No. d–1) Means with different letters within rows are significantly different based on LSD (P < 0.05) yrbooi 20)589:235–248 (2007) Hydrobiologia A/B-ratio of aboveground biomass to belowground (rhizome and root) biomass; S/R-ratio of the biomass of root-supported tissue (aboveground, rhizomes) to root biomass Hydrobiologia (2007) 589:235–248 243

Table 4 Mean (±SE, n = 3) concentrations (g kg–1) of N and P in various plant tissues influenced by nutrient and plant treatments after 65 days of plant growth, with analysis of variance Treatment N concentration P concentration Above-ground Rhizome Root Above-ground Rhizome Root

Low nutrient Mono-C. indica 10.2 ± 0.5de 10.1 ± 1.2c 8.7 ± 0.4cd 3.8 ± 0.1c 3.7 ± 0.1d 2.8 ± 0.1b Mixed-C. indica 10.5 ± 0.1d 9.7 ± 0.3c 8.0 ± 0.4d 4.3 ± 0.1bc 4.0 ± 0.1bcd 2.9 ± 0.1b Mono-S. validus 8.4 ± 0.8e 11.6 ± 1.7c 8.3 ± 0.2d 4.6 ± 0.1ab 3.8 ± 0.1cd 1.4 ± 0.1d Mixed-S. validus 8.4 ± 0.5e 10.9 ± 1.3c 8.7 ± 0.4cd 5.0 ± 0.3a 4.2 ± 0.2abcd 1.6 ± 0.1cd High nutrient Mono-C. indica 20.5 ± 1.2ab 17.1 ± 0.8ab 14.2 ± 0.5ab 4.6 ± 0.1ab 4.3 ± 0.2abc 3.7 ± 0.2a Mixed-C. indica 21.3 ± 0.8a 18.3 ± 0.7a 14.5 ± 0.9a 4.4 ± 0.2abc 4.3 ± 0.2abc 3.4 ± 0.1a Mono-S. validus 19.0 ± 0.4bc 14.9 ± 0.4b 12.7 ± 1.0b 4.6 ± 0.2ab 4.6 ± 0.3a 2.5 ± 0.2b Mixed-S. validus 17.0 ± 0.4c 15.4 ± 0.3ab 10.1 ± 0.4c 4.9 ± 0.3a 4.5 ± 0.1ab 1.9 ± 0.1c Source of variation Nutrient treatment *** *** *** NS *** *** Plant treatment *** NS ** ** NS *** Nutrient · plant NS NS ** NS NS NS Means with different letters within columns are significantly different based on LSD (P < 0.05) NS: not significant. ** Significant at P < 0.01. *** Significant at P < 0.001

Discussion nutrient uptake and use (Tanner, 1996; Gu¨ sewell & Bollens, 2003). In the present study, significant Resource allocation differences in the nutrient use efficiency were not detected among the plant treatments (Table 7). Aquatic plants can take up large quantities of Therefore, the differences in biomass and nutri- nutrients, and even assimilate them luxuriously ent concentration might be related to the nutrient (Cronk & Fennessy, 2001). The present results uptake efficiency. Although the below-ground showed the plants were capable of taking up more biomass was relatively high in the high nutrient N and P (Table 4) and producing more biomass treatment compared with the low nutrient treat- (Table 1) in the high than the low nutrient ment, significant difference in the below-ground treatment. Differences between species in bio- biomass was not observed between the nutrient mass accumulation, and tissue N and P concen- treatments. This might be due to the relatively trations are likely to reflect species and high plant density and small size of the micro- developmental stage differences in efficiency of cosms, which limited below-ground plant growth

Table 5 Analysis of Variable Source of variability variance for nutrient allocation to different Nutrient treatment Plant treatment Interaction tissues of plants between nutrient and plant N/P ratio treatments after 65 days Aboveground ** *** NS of plant growth Rhizome *** *** NS Root NS *** NS N allocation ratio Aboveground (%) *** *** NS Rhizome (%) * *** * Root (%) ** *** NS NS: not significant. P allocation ratio * Significant at P \ 0.05. Aboveground (%) ** *** * ** Significant at Rhizome (%) ** *** * P \ 0.01. *** Significant Root (%) * ** NS at P \ 0.001

123 244 123

Table 6 Mean (±SE, n = 3) N/P ratio, N and P allocation to different tissues of plants influenced by nutrient treatments and plant treatments after 65 days of plant growth Treatment Low nutrient High Nutrient Mono-C. indica Mixed-C. indica Mono-S. validus Mixed-S. validus Mono-C. indica Mixed-C. indica Mono-S. validus Mixed-S. validus

N/P ratio Aboveground 4.8 ± 0.2ab 4.5 ± 0.2ab 3.4 ± 0.2c 3.3 ± 0.2c 4.7 ± 0.1ab 5.1 ± 0.1a 4.4 ± 0.3b 3.6 ± 0.2c Rhizome 2.9 ± 0.1b 2.7 ± 0.1bc 2.3 ± 0.2c 2.1 ± 0.2c 3.5 ± 0.1a 3.5 ± 0.1a 2.9 ± 0.2b 2.3 ± 0.1c Root 3.3 ± 0.1b 3.0 ± 0.03b 7.0 ± 1.0a 6.4 ± 0.6a 2.9 ± 0.3b 3.0 ± 0.1b 5.3 ± 0.9a 6.5 ± 0.9a N allocation ratio Aboveground (%) 45 ± 1d 47 ± 1cd 72 ± 3ab 66 ± 1b 52 ± 1c 56 ± 1bc 75 ± 3a 72 ± 2ab Rhizome (%) 48 ± 1a 47 ± 1a 17 ± 2c 16 ± 1c 43 ± 1ab 40 ± 1b 18 ± 3c 15 ± 1c Root (%) 7 ± 0.2cd 6 ± 1d 11 ± 1bc 18 ± 2a 5 ± 1d 4 ± 0.04d 7 ± 2cd 13 ± 2b P allocation ratio Aboveground (%) 34 ± 2c 35 ± 1c 71 ± 3a 66 ± 1a 44 ± 1b 47 ± 2b 69 ± 3a 70 ± 2a Rhizome (%) 59 ± 2a 58 ± 1a 24 ± 3c 25 ± 0.4c 49 ± 1b 48 ± 2b 26 ± 4c 23 ± 2c Root (%) 7 ± 0.3ab 7 ± 1ab 5 ± 1b 9 ± 1a 7 ± 0.6ab 5 ± 0.2b 5 ± 1b 7 ± 1ab Means with different letters within rows are significantly different based on LSD (P < 0.05) yrbooi 20)589:235–248 (2007) Hydrobiologia Hydrobiologia (2007) 589:235–248 245

Table 7 Mean (±SE, Treatment N use efficiency P use efficiency n = 3) N and P use efficiency (g dry weight –1 Low nutrient g NorP)ofC. indica Mono-C. indica 86 ± 6a 276 ± 5a and S. validus influenced Mixed-C. indica 85 ± 1a 246 ± 3ab by nutrient and plant Mono-S. validus 87 ± 5a 278 ± 4a treatments after 65 days Mixed-S. validus 92 ± 6a 281 ± 21a of plant growth, with High nutrient analysis of variance Mono-C. indica 58 ± 3b 222 ± 10b Mixed-C. indica 56 ± 3b 230 ± 14b Means with different Mono-S. validus 61 ± 3b 243 ± 10b letters within columns are Mixed-S. validus 71 ± 1b 245 ± 13b significantly different Source of variation based on LSD (P \ 0.05) Nutrient treatment *** *** Plant treatment NS NS NS: not significant. *** Nutrient · plant NS NS Significant at P \ 0.001 in the high nutrient treatment. Gu¨ sewell and above-ground to below-ground biomass with Bollens (2003) also pointed out that the total increasing density of Zizania latifolia in mixture. below-ground biomass was less responsive to Resource allocation is known to change during nutrient treatments than the above-ground bio- the growing season for most plant species (Aerts mass, with an inconsistent effect of the N/P supply et al., 1992). The resource allocation patterns ratio in the pot experiments. The total below- observed in the present experiment may only ground biomass was higher in the intermediate reflect the short length of the experiment. It is nutrient supply than the high nutrient supply at 15 important to note that while a microcosm trial N/P supply ratio. enables more control over experimental condi- Plant productivity and resource allocation var- tions than field trials, there are significant differ- ied widely between C. indica and S. validus ences in the temporal and spatial aspects of such (Tables 1, 3 and 6). This variation is likely to studies. Thus the present results of the species arise from relative differences in initial propagule combination, and the relatively short duration of vigor, as well as from intrinsic species and the growth trials in relation to the life cycles of possibly ecotype growth characteristics (Daniels, these clonal species, care must be exercised, in 1991). However, factors such as the physiological attempting to generalize the results from the and developmental state of the propagules are microcosm to field-scale constructed wetland. likely to have been of more importance than the biomass of the propagules per se (Tanner, 1996). Competition in mixture The A/B ratios of the species in the present study ranged from 0.6 to 1.7. These values were in Grime and Hodgson (1987) listed characteristics the ranges of other aquatic plants (Hogetu, 1984). of species with high competitive ability: (1) a Plants alter their resource allocation to above- robust perennial life form with a strong capacity ground and below-ground tissues along environ- to ramify vegetatively, (2) the rapid commit- mental gradients of disturbance and resource ment of captured resources to the construction availability (Kirkman & Sharitz, 1993; Stuart of new leaves and roots, (3) high morphological et al., 1999). In the present study, the resource plasticity during the differentiation of leaves allocations were altered by nutrient availability, and roots, and (4) short life spans of individual and also changed for S. validus by the mixture. leaves and roots. Both C. indica and S. validus The growth of S. validus in mixture was strongly are robust perennials, which rapidly produce inhibited by the presence of C. indica, and allo- ramets and have high growth rates. In addition, cated more percentage biomass to its below- C. indica produces large storage rhizomes and ground tissues. Wu & Yu (2004) found that high growth rates have been measured (Zhao Nymphoides peltata decreased the ratios of et al., 2003).

123 246 Hydrobiologia (2007) 589:235–248

Wetzel and van der Valk (1998) pointed out with monoculture of C. indica, followed by mix- that rapid growth was not the only factor, and ture, but no significant difference was found suggested that plant architecture played a signif- between monoculture of S. validus and the treat- icant role in competition between Carex stricta, ment without plants (see Zhang et al., 2007). Phalaris arundinacea and Typha latifolia. Mor- phological characteristic of a plant affecting competition for light have been reported in Conclusions agricultural and woody plants (McLachlan et al., 1993; Sipe & Bazzaz, 1994; Webster et al., 1994). High nutrient availability improved plant growth Species with different morphologies showed large and resulted in allocation of more resources to differences in canopy structure. A grass, having a the above-ground tissues compared to below- more open canopy, was consistently a weak ground ones. Due to interspecific competition, the competitor when grown with forbs (Tremmel & growth and resource allocation of S. validus were Bazzaz, 1993). The morphological characteristics, significantly influenced by mixture, but C. indica such as tall shoot, shape and large canopy was less affected. The results suggested that the diameter, were significantly correlated with in- plant growth and resource allocation of C. indica creased competitive ability in wetland plants and S. validus could be altered by nutrient avail- (Gaudet & Keddy, 1988). Changes of water ability and interspecific competition in the con- levels, and the presence/absence of competitor structed wetlands. To enhance the aesthetic species, produced significant morphological re- appeal of constructed wetlands, but avoid the sponses in mature individuals of five freshwater interspecific competition, the intensive studies on wetland plant species: Agrostis stolonifera L., nutrient uptake for each plant species and their Carex rostrata Stokes, Deschampsia cespitosa mixtures at various nutrient concentrations and at (L.) Beauv., Filipendula ulmaria (L.), Phalaris various planting densities are needed in both arundinacea (L.). These responses provide evi- laboratory and field conditions. dence for potential advantages in survival and ability to spread vegetatively (Kennedy et al., 2003). In the present study, C. indica and S. val- References idus differ substantially in their growth rate, morphology, physiology and size. It is possible Aerts, R. & F. S. Chapin, 2000. The mineral nutrition of that C. indica, having large leaf areas and canopy wild plants revisited: a re-evaluation of processes and patterns. Advances in Ecological Research 30: diameter, maximized the capture of light and 1–67. nutrient resources by maximizing vegetative Aerts, R., H. de Caluwe & H. Konings, 1992. Seasonal growth under both nutrient availabilities and allocation of biomass and nitrogen in four Carex out-competed S. validus in mixture. species from mesotrophic and eutrophic fens as af- fected by nitrogen supply. Journal of Ecology 80: 653– Interspecific competition is often regarded as 664. being caused by mutual exploitation of limiting Agami, M. & K. R. Reddy, 1990. Competition for space resources (resource consumption, including light between Eichhornia crassipes (Mart.) Solms and Pis- interception by plants and space occupancy by tia stratiotes L. cultured in nutrient-enriched water Aquatic Botany 38: 195–208. space-limited sessile organisms), by the production Bassett, J., R. C. Denney, G. H. Jeffery & J. Mendham, of toxins, and by various combinations of these 1978. Vogel’s Textbook of Quantitative Inorganic mechanisms (Tilman, 1987). In the present results, Analysis Including Elementary Instrumental Analysis. however, it is unknown whether interspecific com- 4th edition. Longman, London, New York. Calheiros, C. S. C., A. O. S. S. Rangel & P. M. L. Castro, petition between C. indica and S. validus was 2007. Constructed wetland systems vegetated with caused by single mechanism or the combinations different plants applied to the treatment of tannery of above mentioned mechanisms. It is worth wastewater. Water Research DOI 10.1016/j.watres. mentioning that there were significant differences 2007.01.012. Coleman, J., K. Hench, K. Garbutt, A. Sextone, G. in pH of the effluents among the plant treatments. Bissonnette & J. Skousen, 2001. Treatment of The lowest effluent pH was detected in microcosms domestic wastewater by three wetland plant species in

123 Hydrobiologia (2007) 589:235–248 247

constructed wetlands. Water, Air and Soil Pollution tive review. Australian Journal of Plant Physiology 27: 128: 283–295. 595–607. Cronk, J. K. & M. S. Fennessy, 2001. Wetland plants: Shi, L., B. Wang, X. Cao, J. Wang, Z. Lei, Z. Wang, Z. Liu Biology and Ecology. Lewis Publisher, Boca Raton, & B. Lu, 2004. Performance of a subsurface-flow FL, USA, 462 pp. constructed wetland in southern . Journal of Daniels, R. E., 1991. Variation in the performance of Environmental Sciences 16: 476–481. Phragmites australis in experimental culture. Aquatic Sipe, T. W. & F. A. Bazzaz, 1994. Gap partitioning among Botany 42: 41–48. Maples (Acer) in central New England: shoot archi- EPA, 2000. U.S. Environmental Protection Agency, tecture and photosynthesis. Ecology 75: 2318–2332. Manual, Constructed Wetlands Treatment of Muni- Stuart, J. B., G. G. Georage & K. F. Walker, 1999. Growth cipal Wastewaters. EPA/625/R-99/010. Cincinnati, and resource allocation in response to flooding in the OH. emerged sedge Bolboschoenus medianus. Aquatic Fu, C., Y. Tang, X. Chen, Y. Xu & G. Xing, 2006. Effects Botany 63: 145–160. of purification highly salty reuse water quality by Sundaravadivel, M. & S. Vigneswaran, 2001. Constructed three plants in TEDA landscape river. Journal of wetlands for wastewater treatment. CRC Critical Chongqing University (Natural Science Edition) 29: Reviews in Environmental Science and Technology 118–121 (in Chinese, with English abstract). 31: 351–409. Gaudet, C. L. & P. A. Keddy, 1988. A comparative ap- Tanner, C. C., 1996. Plants for constructed wetland treat- proach to predicting competitive ability from plant ment systems–a comparison of the growth and nutri- traits. Nature 334: 242–243. ent uptake of eight emergent species. Ecology Grime, J. P. & J. G. Hodgson, 1987. Botanical contribu- Engineering 7: 59–83. tions to contemporary ecological theory. New Phy- Tanner, C. C., 2001. Growth and nutrient dynamics of soft- tologist 106(Suppl.), 283–295. stem bulrush in constructed wetland treating nutrient- Grosse, W., F.W. Wissing, R. Perfler, Z. Wu, J. Chang & Z. rich wastewaters. Wetland Ecology and Management Lei, 2001. Water Quality Improvement in Tropical 9: 49–73. and Subtropical Areas for Reuse and Rehabilitation Tilman, D., 1987. The importance of the mechanisms of of Aquatic Ecosystem. In Gawlik, B.M., B. Platzer & interspecific competition. The American Naturalist H. Muntau (eds), Freshwater Contamination in Chi- 129: 769–774. na. Office for official publications of the European Tremmel, D. C. & F. A. Bazzaz, 1993. How neighbour Communities, European Communities, 19–32. canopy architecture affects target plant performance. Gu¨ sewell, S. & U. Bollens, 2003. Composition of plant Ecology 74: 2114–2124. species mixtures grown at various N:P ratios and Tylova-Munzarova, E., B. Lorenzen, H. Brix & O. Vot- + – levels of nutrient supply. Basic and Applied Ecology rubova, 2005. The effects of NH4 and NO3 on growth, 4: 453–466. resource allocation and nitrogen uptake kinetics of Harper, J. L., 1977. Population Biology of Plants. Aca- Phragmites australis and Glyceria maxima. Aquatic demic Press, New York. Botany 81: 326–342. Hogetu, K., 1984. Bioeconomics. Syokabo Press. Tokyo, Webster, T. M., M. M. Loux, E. E. Regnier & S. K. Har- 245. rison, 1994. Giant ragweed (Ambrosia trifida) canopy Kadlec, R. H. & R. L. Knight, 1996. Treatment Wetlands. architecture and interference studies in Soybean Lewis Publisher, Boca Raton, FL, USA, 893 pp. (Glycine max). Weed Technology 8: 559–564. Kennedy, M. P., J. M. Milne & K. J. Murphy, 2003. Wetzel, P. R. & A. G. van der Valk, 1998. Effect of Growth responses to groundwater level variation and nutrient and soil moisture on competition between competition in freshwater wetland plant species. Phalaris arundinacea, Carex stricta and Typha latifo- Wetlands Ecology and Management 11: 383–396. lia. Plant Ecology 138: 179–190. Kirkman, L. K. & R. R. Sharitz, 1993. Growth in con- Wu, J. & L. Ding, 2006. Study on treatment of polluted trolled water regimes of three grasses common in river water using pilot-scale surface flow constructed freshwater wetlands of the southeastern USA. wetlands system. Environmental Pollution and Con- Aquatic Botany 44: 345–359. trol 28: 432–434 (in Chinese, with English abstract). Lorenzen, B., H. Brix, I. A. Mendelssohn, K. L. McKee & Wu, J., S. Huang, X. Ruan & L. Ding, 2006. Treatment of S. L. Miao, 2001. Growth, biomass allocation and polluted river water using surface flow constructed nutrient use efficiency in Cladium jamaicense and wetlands in Xinyi river floodplain, Jiangsu Province. Typha domingensis as affected by phosphorus Journal of Lake Sciences 18: 238–242 (in Chinese, and oxygen availability. Aquatic Botany 70: 117–133. with English abstract). McLachlan, S. M., M. Tollenaar, C. J. Swanton & S. F. Wu, Z. & D. Yu, 2004. The effects of competition on Weise, 1993. Effect of corn-induced shading on dry growth and biomass allocation in Nymphoides peltata matter accumulation, distribution, and architecture of (Gmel.) O. Kuntze growing in microcosm. Hydrobi- redroot pigweed (Amaranthus retroflexus). Weed ologia 527: 241–250. Science 41: 568–573. Yue, C. L., J. Chang, Y. Ge, & Y. M. Zhu, 2004. Treatment Poorter, H. & O. Nagel, 2000. The role of biomass allo- efficiency of domestic wastewater by vertical/reverse- cation in the growth response of plants to different vertical flow constructed wetlands. Fresenius Envi- levels of light, CO2, Nutrients and water: a quantita- ronmental Bulletin 13: 505–507.

123 248 Hydrobiologia (2007) 589:235–248

Zhang, Z., Z. Rengel & K. Meney, 2007. Nutrient removal wetland. China Environmental Science 23: 290–294 from simulated wastewater using Canna indica and (in Chinese, with English abstract). Schoenoplectus validus in mono- and mixed-culture in Zhu, X., L. Cui, W. Liu & Y. Liu, 2004. Removal effi- wetland microcosms. Water, Air and Soil Pollution ciencies of septic tank effluent by simulating vertical- DOI 10.1007/s11270–007–9359–3. flow constructed Canna indica Linn wetlands. Journal Zhao, J., Q. Yang, Z. Chen & Z. Huang, 2003. Studies on of Agro-Environment Science 23: 761–765 (in Chi- root system biomass of the plants in several kinds of nese, with English abstract).

123